Podcast transcript generator

Upload an episode or an interview and get a transcript split by speaker — who said what, and when. Give the speakers their names and download it. Free, no sign-up.

How it works

  1. Upload the episode

    An MP3 or M4A export, a WAV from your recorder, or the video of a video podcast. Your browser takes only the sound and uploads that. Up to 30 minutes per file — split longer episodes into parts.

  2. Wait a minute or two

    The speech is transcribed and the voices told apart. The first few minutes decide who is who; the rest of the episode is matched against those voices, so "Speaker 1" stays the same person from start to end.

  3. Name the speakers and download

    Type the real names over Speaker 1, Speaker 2 — the transcript updates as you type. Copy it, or download TXT for show notes, SRT with the names in each line, or VTT with proper voice tags.

What the transcript looks like

Each turn starts with the speaker and the time it began:

Anna (0:00)
Welcome back to the show. Today I'm talking with Daniel about building an app on his own.

Daniel (0:07)
Thanks for having me. It's been a long year.

Getting the speakers right

Separate microphones and little crosstalk give the cleanest result. Voices that sound very alike, people talking over each other, or a guest who only says a word or two can end up with the wrong label — skim the transcript before you publish it. Up to four voices are tracked across the whole episode; a fifth person is labelled where they're heard.

Show notes, quotes and clips

Timestamps make it easy to find the moment a quote came from. For short vertical clips of the episode, the Subtitled app on iPhone adds captions that pop word by word in time with the voice — see word-by-word captions.

Is the recording kept?

No. The audio is sent to our server only to be transcribed and labelled by OpenAI, which doesn't train on it, and is deleted as soon as the transcript is ready. More in the privacy policy.

Get Subtitled for iPhone